Map matching and uncertainty: an algorithm and real-world experiments

Authors:
Kristof Ghys;Bart Kuijpers;Bart Moelans;Walied Othman;Dries Vangoidsenhoven;Alejandro Vaisman
Affiliations:
Hasselt University, Belgium;Hasselt University, Belgium;Hasselt University, Belgium;Hasselt University, Belgium;Hasselt University, Belgium;Universidad de Buenos Aires, Argentina
Venue:
Proceedings of the 17th ACM SIGSPATIAL International Conference on Advances in Geographic Information Systems
Year:
2009

Citing 4
Cited 1

Finding the Lengths of All Shortest paths in N -Node Nonnegative-Distance Complete Networks Using 12N3 Additions and N3 Comparisons

Journal of the ACM (JACM)
Modeling Moving Objects over Multiple Granularities

Annals of Mathematics and Artificial Intelligence
Capturing the Uncertainty of Moving-Object Representations

SSD '99 Proceedings of the 6th International Symposium on Advances in Spatial Databases
Mobility, Data Mining and Privacy: Geographic Knowledge Discovery

Mobility, Data Mining and Privacy: Geographic Knowledge Discovery

Map-based spatio-temporal interpolation in vehicle trajectory data using routing web-services

Proceedings of the 5th ACM SIGSPATIAL International Workshop on Computational Transportation Science

Quantified Score

Hi-index	0.00

Visualization

Abstract

A common problem in moving object databases (MOD) is the reconstruction of a trajectory from a trajectory sample (i.e., a finite sequence of time-space points). A typical solution to this problem is linear interpolation. A more realistic model is based on the notion of uncertainty modelled by space-time prisms, which capture the positions where the object could have been, when it moved from a to b. Often, object positions measured by location-aware devices are not on a road network. Thus, matching the user's position to a location on the digital map is required. This problem is called map matching. In this paper we study the relation between map matching and uncertainty, and propose an algorithm that combines weighted k-shortest paths with space-time prisms. We apply this algorithm to two real-world case studies and we show that accounting for uncertainty leads to obtaining more positive matchings.